基于神经网络的页岩微纳米孔隙微结构分析的正则化和最优化方法  

Regularization and optimization methods for micro pore structure analysis of shale based on neural networks

作　　者：王彦飞[1,2,3] 邹安祺[1,2,3] 

机构地区：[1]中国科学院地质与地球物理研究所,中国科学院油气资源研究重点实验室,北京100029 [2]中国科学院地球科学研究院,北京100029 [3]中国科学院大学,北京100049

出　　处：《岩石学报》2018年第2期281-288,共8页Acta Petrologica Sinica

基　　金：本文受中国科学院先导科技专项(XDB10020100)和国家自然科学基金项目(91630202、41611530693、41325016)联合资助.感谢中国科学院先导科技专项(B类)提供页岩样品和上海光源提供同步辐射数据.感谢审稿专家的宝贵建议.

摘　　要：页岩气成藏机理与页岩内部孔隙结构紧密相关,对页岩孔隙结构的研究成为页岩气勘探开发技术中至关重要的一环。页岩内部不同结构体组分对X-射线的吸收能谱不一样,这样就导致观测数据是由不同页岩组分衰减不同波段的X-射线构成的。经过对CT图像分割,能够获得页岩微孔结构的图像,尤其是获得有机质中孔隙类别、形状、尺寸、空间分布、连通特性。本文利用同步辐射X射线扫描重构的页岩CT数据,研究并设计基于多能CT图像的神经网络图像分割技术和算法,以期得到页岩体三维结构特征及空间分布,可以为建立有机质种类和无机矿物组成与微纳孔隙特征的联系以及最终实现页岩气的资源储量评估和勘探开发提供技术支持。The mechanism of shale gas accumulation is closely related to the internal pore structure of shale. The study of the shale pore structure has become an important issue of the shale gas exploration technology. Since different structural components within the shale possess different X-ray absorption spectra, the observed CT data consists of different shale components attenuating X-rays in different bands. With the CT image segmentation, the images of shale micropores can be obtained, especially the pore type, shape, size, spatial distribution and connectivity. In this paper, based on the reconstructed synchrotron radiation X-ray shale CT data, we develop a neural network image segmentation technology and algorithm based on multi-energy CT image data in order to obtain the 3D structural characteristics and spatial distribution of shale. The new technology can be used to establish the relationship between the organic matter species and the inorganic mineral composition, so as to obtain the micro-and nano-pore features, as well as to provide technical support for the assessment of reserves of the shale gas resource and for the exploration.

关 键 词：页岩微米孔隙结构 CT图像分割 神经网络计算 最优化与正则化 

分 类 号：P313.1[天文地球—固体地球物理学] P628[天文地球—地球物理学]